Management of aortic diseases has progressed dramatically since the first successful, reproducible surgical intervention in 1956; however, although our understanding of the genetic and cellular bases of such diseases has steadily grown, treatment planning still generally relies on simple risk-assessment models and clinical experience. Some pathological conditions have been mimicked with animal models, but results from such studies may not be readily extrapolated to patients. Other pathologies lack any accepted or reproducible animal model. An example is aortic dissection, in which an intimal tear in the aortic wall propagates into the media to form a false lumen within the vessel wall. Surgical treatment for aortic dissection consists of either replacement of a portion of the aorta, or endovascular stent implantation to cover the affected segment. Both approaches carry significant risks, and determining the optimal choice and timing of an intervention is challenging. Because there are no accepted animal models of aortic dissection, experimental studies must use physical or computational models. Existing computational models of aortic dissection use conventional computational fluid dynamics (CFD) approaches, in which the vessel wall and flap are treated as rigid structures. Although CFD models are able to predict wall shear stress distributions, they are unable to account for the interactions between the blood and vascular tissues, or for the effects of such interactions on the dynamics of the dissected aorta. This project will develop fluid-structure interaction (FSI) models of aortic dissection that overcome th limitations of CFD models. These predictive models will be used to perform patient-specific simulations that ultimately will aid in clinical decision making, e.g., selecting optimal medical therapies or surgical interventions. This project will develop two types of FSI models of aortic dissection. The first type of model will use an idealized description of the geometry of the vessel and lesion. Such models are ideally suited for addressing questions that take the form of parameter studies. These models will be used to study systematically how geometry and driving conditions affect the dynamics of both developing dissections and fully developed lesions. The second type of model will account for the effects of patient-specific anatomy by using geometries derived from computed tomography (CT) and/or magnetic resonance (MR) imaging studies. To characterize the elastic response of human aortic tissue, tissue samples will be collected from both normal and diseased human aortas, and tensile tests will be performed to characterize the mechanical properties of these specimens. The data from these tests will be used to develop corresponding healthy and disease-specific constitutive models. The characterization of the elasticity of both the healthy and diseased human aorta has the potential to impact work on a broad range of aortic diseases. Finally, these models will be used to study the medical and surgical management of patients who require or who have undergone only partial surgical repair of the dissection, as is now commonly done in cases in which the dissection involves the proximal ascending aorta (Type A dissections).